Efficient estimation of the nonlinear aerodynamic loads of floating offshore wind turbines under random waves and wind in frequency domain

This paper investigates an efficient approach to estimate nonlinear wind turbine loads of floating offshore wind turbines in the frequency domain. The international benchmark of the 5-MW NREL reference wind turbine supported by the OC3-Hywind Spar platform is chosen as a case study. The modelling of the floating offshore wind turbine subjected to random wind and wave forces is given in the frequency domain. The present formulation captures the effects of the wind turbulence acting on the platform which leads to low-frequency motions. To improve the aerodynamic load estimation, the statistical linearisation (SL) technique is used to replace the nonlinear aerodynamic functions by their equivalent linear ones, where the differences are minimised in a mean squared sense using an iterative procedure. The SL technique is compared against its equivalent nonlinear time-domain model (for benchmark) and other conventional frequency-domain models. The results demonstrate good agreement at the maximum power control region for all frequency domain-based models, whilst only SL provided accurate estimates of the response at the rated power control region. This strong agreement using SL is demonstrated in terms of probability distributions, power spectrum densities, and mean results for the displacements, collective blade pitch angle, and rotational speed of the generator. Despite its accurate estimations, the main advantage of SL is the low computational cost, which is approximately two to three orders of magnitude faster than the equivalent nonlinear time-domain simulations. Hence, SL can be a useful approach for reliable preliminary investigations, optimisation procedures, and design of the wind turbine controller.